Background: Clinical decision support (CDS) tools in oncology have been shown to improve care quality, reduce care variation and reduce healthcare costs. Flatiron Assist® (FA) is a customizable CDS tool that is embedded in the electronic health record (EHR) to facilitate selection and documentation of National Comprehensive Cancer Network (NCCN) guideline concordant treatment regimens. Our objective was to compare the distribution of regimens ordered before and after implementation of FA at a community cancer clinic to determine if variation in regimens ordered was impacted by implementing the CDS tool. Methods: This study used the nationwide EHR-derived de-identified Flatiron Health database to analyze treatment regimens ordered for Breast, Colon, Non–Small Cell Lung, Rectal, Prostate and Small Cell Lung Cancers placed at a multi-site community cancer clinic over a 58 month observation period (11/14/2017 to 9/14/2022). FA was implemented on 5/19/20. We compared the total number of unique regimens with respect to the total number of orders placed before and after implementation. We calculated the proportion reduction in the number of regimens ordered with Wald (normal approximation) 95% confidence intervals to describe the relationship between the regimen variation seen before and after implementation of FA. Results: A total of 9,530 orders placed by 57 prescribers were included in the analysis. Before implementation, 5,112 orders were placed consisting of 1,817 unique regimens. After implementation, 4,418 orders were placed consisting of 1,018 unique regimens. We stratified the results to assess reduction in variation by comparing the number of regimens required to reach 75, 80, 90 and 100% of orders respectively. (Table 1). Overall, the number of regimens to reach 100% of orders was reduced by 44.0% (95% confidence interval: 41.7–46.3%, p<0.01) after implementing FA. Conclusions: Following adoption of FA, there was a significant reduction in the number of unique regimens with respect to total regimens ordered. A limitation of this study includes the inability to account for additional factors that may contribute to this reduction in variation. Regardless, these findings suggest that implementing an EHR-embedded CDS tool contributes to a reduction in care variation by standardizing unique regimens ordered at a multi-site community cancer clinic. Future studies will explore other quality metrics that may be impacted by implementing FA.
